# Prompt

> Generate, rewrite or summarize text with Gemini, Claude, GPT or Grok. Add instructions, images, video or audio, and split the answer into items.

Source: https://nodaro.ai/docs/nodes/automate/prompt

The **Prompt** node sends your text to an AI language model and returns the model's answer as text. Use it to write, rewrite, summarize, translate or describe. Choose a model, write instructions and a request, and wire the answer into any node that accepts text. The node was called **Generate Text** before September 2026. Workflows that contain the older **AI Agent** node (type `ai-writer`) open with a Prompt node in its place.

- Found in: Automate › Text
- Output: text
- API type: `llm-chat`

## When to use it
- Write or rewrite text: captions, scripts, product descriptions, headlines, translations.
- Turn one idea into several image prompts, and generate one image per prompt.
- Describe or summarize a connected image, video or audio clip.
- Summarize or rewrite content from [Web Scrape](https://nodaro.ai/docs/nodes/automate/web-scrape), [Telegram Channel Feed](https://nodaro.ai/docs/nodes/automate/telegram-channel-feed) or [Transcribe](https://nodaro.ai/docs/nodes/audio/transcribe).
- Return structured data, such as JSON, for [Extract Field](https://nodaro.ai/docs/nodes/automate/extract-field).

## Quick start
### Add the node

Press Tab on the canvas and choose **Automate › Text › Prompt**. The Prompt node is also on the **Common** tab.

### Choose a model

Open the settings panel and choose a model under **AI Model**. The default, **Gemini 3.6 Flash**, costs 1 credit per run and suits most rewriting and captioning.

### Write the instructions and the request

Under **Instructions (System Prompt)**, say how the model should answer: its role, the tone, the length, the format. Under **User Prompt**, write the request itself. You can also wire a [Text](https://nodaro.ai/docs/nodes/automate/text) node into the **Prompt** input instead.

### Run it

Click **Run** on the node. The answer appears on the node while the model writes it, and every earlier answer stays available on the node.

Workflow: A Text node gives the idea, the Prompt node writes four image prompts on its Items output, and Generate Image runs once for each.

- Text → Prompt (prompt)
- Prompt → Generate Image (prompt)

![The Prompt node settings panel with the AI Model, Reasoning Effort and Advanced mode controls above the Instructions and User Prompt fields.](https://nodaro.ai/docs-media/screens/en/nodes/prompt/settings.light.webp)

## Inputs
The Prompt node has three inputs on its left edge.

| Input | Accepts | What it does |
| --- | --- | --- |
| **Prompt** | Text nodes, such as Text, Combine Text and Split Text, pickers, and the JSON of Web Scrape | The request. It is used when **User Prompt** is empty. |
| **References** | Image, video, audio and text nodes | Media for the model to look at or listen to, for example "describe this image". A connected text is added to the prompt as extra context. Video and audio need a model that accepts them, see [Models](#models). |
| **Instructions** | Text nodes | The instructions. They are used when **Instructions (System Prompt)** is empty. |

When a field has typed text and its input is also connected, the typed text wins. To use both, write `{}` in the typed text where the text of the **Prompt** input should go.

The node has two outputs:

| Output | Carries | Use it to |
| --- | --- | --- |
| **Text** | The whole answer as one text. Any `===NEXT===` markers stay in it. | Feed the whole answer to Combine Text, a publishing node or a second Prompt node. |
| **Items** | The answer split at every `===NEXT===` marker, one item per part. With no marker, it holds one item. | Run the next node once per item, for example one image per prompt. |

## Settings
| Setting | What it does |
| --- | --- |
| **Preset** | **Custom** (the default) for your own instructions, a built-in preset, or one of your saved presets. See [Presets](#presets). |
| **AI Model** | The language model. The default is **Gemini 3.6 Flash**. Search the list by name, company or tier. |
| **Reasoning Effort** | How much the model reasons before it answers: **Auto (model default)**, **None — no reasoning**, **Low**, **Medium**, **High**, **Very high (may bill one tier up)** or **Max (may bill one tier up)**. Each model offers its own levels, and the control is hidden for models that have none. |
| **Advanced mode** | Gemini models only. Runs the model directly with its maker, so that **Temperature**, **Max Tokens** and every reasoning level take effect. Costs one credit tier more. |
| **Instructions (System Prompt)** | Optional instructions that guide the answer: role, tone, length, format and rules. |
| **User Prompt** | The request. It can contain variables, such as `{Product}`. |
| **Temperature** | From 0 to 2. The default is 0.7. Lower values give steadier, more factual answers, and higher values give more varied ones. Some models ignore it, see [Models](#models). |
| **Max Tokens** | The longest answer the model may write, in tokens. The default is 8,192. See [When the answer is cut off](#when-the-answer-is-cut-off). |
| **References** | Media you add by public URL instead of by connection: up to 5 images, and on models that accept them up to 3 videos and 3 audio clips. |
| **Number of runs** | How many answers one click of **Run** produces, from 1 to 4, in the toolbar under the node. Each answer is charged. |
| **Pre & post text** | Text that is always added before and after the request, never to the instructions. It is hidden from people who use your workflow as an app. See [Prompt pre and post text](https://nodaro.ai/docs/concepts/prompt-pre-post-text). |

The toolbar under the node repeats the most used settings: the model, the reasoning effort, the preset and the number of runs.

## Models
The Prompt node can run the language models below. The **AI Model** picker in the settings panel always shows the current list.

| Model | Maker | Tier | Credits per run | References it reads |
| --- | --- | --- | --- | --- |
| Gemini 3 Flash | Google | Economy | 1 | Image, video, audio |
| Gemini 3.6 Flash (default) | Google | Economy | 1 | Image, video, audio |
| Gemini 3.7 Flash | Google | Economy | 1 | Image |
| Gemini 3.8 Flash | Google | Economy | 1 | Image |
| Claude Haiku 4.5 | Anthropic | Economy | 1 | Image |
| GPT-5.6 Luna | OpenAI | Economy | 1 | Image |
| Claude Sonnet 4.6 | Anthropic | Standard | 2 | Image |
| Claude Sonnet 5 | Anthropic | Standard | 2 | Image |
| GPT-5.2 | OpenAI | Standard | 2 | Image |
| GPT-5.6 Terra | OpenAI | Standard | 2 | Image |
| Grok 4.6 | xAI | Standard | 2 | Image |
| Gemini 3.1 Pro | Google | Premium | 6 | Image, video, audio |
| Claude Opus 4.7 | Anthropic | Premium | 6 | Image |
| Claude Opus 4.8 | Anthropic | Premium | 6 | Image |
| Claude Opus 5 | Anthropic | Premium | 6 | Image |
| Claude Fable 5 | Anthropic | Premium | 6 | Image |
| GPT-5.4 | OpenAI | Premium | 6 | Image |
| GPT-5.5 | OpenAI | Premium | 6 | Image |
| GPT-5.6 Sol | OpenAI | Premium | 6 | Image |
| GPT-6 Astra | OpenAI | Premium | 6 | Image |

- **Video and audio.** Only Gemini 3 Flash, Gemini 3.6 Flash and Gemini 3.1 Pro read video and audio references. With any other model, a connected video or audio reference is ignored.
- **Temperature.** These models ignore **Temperature**: Claude Opus 4.7, Claude Opus 4.8, Claude Opus 5, Claude Sonnet 5, Claude Fable 5, GPT-5.5, GPT-5.6 Luna, GPT-5.6 Terra, GPT-5.6 Sol, GPT-6 Astra and Grok 4.6.
- **Reasoning Effort.** Gemini 3 Flash, Gemini 3.1 Pro, Claude Haiku 4.5 and GPT-5.2 have no effort levels. The two Gemini models gain levels in **Advanced mode**. Gemini 3.6 Flash, Gemini 3.7 Flash and Gemini 3.8 Flash offer **Low** and **High**, and more levels in Advanced mode.

### Which model to choose

- **Economy models** are plenty for rewriting, captions, translations and image descriptions. Start with the default, Gemini 3.6 Flash.
- **Standard and premium models** are worth their price for long, complex writing and for reasoning over many rules.
- **For a video or audio clip**, choose Gemini 3 Flash, Gemini 3.6 Flash or Gemini 3.1 Pro.

Choose the cheapest model that meets your quality bar, and compare two models on the same request before a large batch.

## Credits
The price of a run depends on the tier of the model:

| Tier | Credits per run |
| --- | --- |
| Economy | 1 |
| Standard | 2 |
| Premium | 6 |

- **Very high and Max bill one tier up.** An economy model bills 2 credits, and a standard model bills 6. A premium model stays at 6. **High** and the lower levels never change the price.
- **Advanced mode bills one tier up.** A Gemini Flash model bills 2 credits in Advanced mode. Gemini 3.1 Pro stays at 6.
- **Number of runs multiplies the price.** Four runs of a standard model cost 8 credits, and the **Run** button shows the total.
- **A failed run is refunded,** including an answer that was cut off.

For example, GPT-5.6 Luna at **Max** costs 2 credits, GPT-5.6 Terra at **Very high** costs 6 credits, and Claude Sonnet 5 at **High** costs 2 credits.

## Split the answer into items

The Prompt node can turn one request into many downstream runs. Ask the model to separate the parts of its answer with a line that reads `===NEXT===`, for example: "Write 4 image prompts for this product. Separate them with ===NEXT===." The **Items** output splits the answer at each marker, and a node connected to **Items** runs once per item.

After a run, the settings panel also lists the items under **Generated Prompts**, with two buttons for working on the canvas by hand:

- **Create N Image Nodes** adds one [Generate Image](https://nodaro.ai/docs/nodes/image/generate-image) node per item, already connected and filled with its prompt.
- **Generate All N Images** runs every node that the first button created.

Use these buttons while you explore. For a workflow that runs on its own, connect the **Items** output instead.

### Presets

A preset sets the instructions and the output format.

| Preset | What it writes | Needs a reference image |
| --- | --- | --- |
| **Custom** | Anything: you write the instructions yourself | No |
| **Photo Shoot Planner** | A series of image prompts for a photo shoot, with scenes, poses, lighting and wardrobe | Yes |
| **Product Catalog Writer** | Image prompts for e-commerce product photos | Yes |
| **Storyboard Writer** | Scene-by-scene image prompts for a video | Yes |

The three writers separate their prompts with `===NEXT===`, ready for the **Items** output. They need a reference image, such as an [Upload Image](https://nodaro.ai/docs/nodes/image/upload-image) or a Generate Image node, connected to **References**. Without one, the panel shows a warning and the node refuses to run.

To keep your own instructions, click **Save as preset**. When one of your presets is selected, **Update** saves your changes into it, and the delete button removes it. Your presets also appear in **Settings**. Read more in [Presets](https://nodaro.ai/docs/concepts/presets).

## Use variables in the prompt

You can put the value of another node in **User Prompt** or **Instructions (System Prompt)** by writing its label in curly braces. `{Product}` becomes the text of a node labeled Product, and `{Mood}` becomes whatever the Mood picker is set to.

- **A default value.** `{tone || friendly}` uses the connected node labeled tone when there is one, and the word "friendly" when there is not.
- **The connected input.** `{}` inserts the text that arrives on the **Prompt** input at that point.
- **One run per item.** When a list, such as the output of [Split Text](https://nodaro.ai/docs/nodes/automate/split-text), arrives on the **Prompt** input, the node runs once per item. Each item replaces **User Prompt**. Put the task in **Instructions (System Prompt)**, which applies to every item. **Pre & post text** also wraps every item.
- **Highlighting.** A variable is cyan when a matching node exists and amber when nothing provides it yet. Amber is only a warning: a variable with a default still works.
- **A missing node stops the run.** If a variable names a node that does not exist anywhere in the workflow, the run is refused before any credits are spent.

Read more in [Prompt variables](https://nodaro.ai/docs/concepts/prompt-variables).

## Ask for structured output

To use the answer as data, ask for a format in the instructions. For example, write "Answer only with a JSON object with the keys title, caption and hashtags". A low **Temperature** keeps the format steadier on the models that accept it.

On the node, the view button switches the answer between raw text and a rendered view. **View as JSON** shows valid JSON as a colored object, and **View as Markdown** renders any other answer. Connect the **Text** output to [Extract Field](https://nodaro.ai/docs/nodes/automate/extract-field) to pull one field, such as `caption`, out of the JSON.

## Read and reuse the result

The toolbar at the top of the result on the node has:

- The view button, to switch between raw text and the rendered view.
- **Copy text** and **Download**.
- **Open log**, the full log of every run, with the model and the preset of each.
- **Show outputs**, when the node has more than one result: a strip of numbered results to page through. The selected result is the one that the outputs pass on.

Each result names the model and the preset that produced it. Drag the magnifier at the bottom-left corner of the node to enlarge it for reading.

## When the answer is cut off

If the model reaches **Max Tokens** before it finishes, the run fails with a "cut off" error and the credits are refunded. A partial answer is never saved or passed on. To fix it, ask for a shorter answer in the instructions, lower **Reasoning Effort**, or raise **Max Tokens**.

Some models always reason before they answer, and their reasoning counts toward **Max Tokens**. On these models, Nodaro raises a small **Max Tokens** value to a minimum, so that the answer keeps room:

| Models | Minimum Max Tokens |
| --- | --- |
| Claude Opus 5, Grok 4.6, GPT-6 Astra | 32,768 |
| Gemini 3.8 Flash, Gemini 3.1 Pro | 16,384 |
| Gemini 3 Flash, Gemini 3.6 Flash, Gemini 3.7 Flash | 8,192 |

Nodaro also raises a small **Max Tokens** value on any model set to **Very high** or **Max**.

## Tips
- **Put the rules in the instructions.** Format, tone, length and "never" rules belong in **Instructions (System Prompt)**, and the request belongs in **User Prompt**. Good instructions give more consistent answers.
- **Adjust Temperature to the task.** Keep 0.7 for most text, lower it for facts and structured output, and raise it for brainstorming.
- **Compare answers side by side.** Set **Number of runs** to 2 or more, then page through the results with **Show outputs**.
- **Write reusable prompts once.** Save instructions you use often with **Save as preset**, or keep a house style in **Pre & post text**.

## Frequently asked questions

### Which AI models can the Prompt node use?

Twenty language models from Google, Anthropic, OpenAI and xAI, including Gemini 3.6 Flash (the default), Claude Sonnet 5, Claude Opus 5, GPT-5.6 Terra, GPT-6 Astra and Grok 4.6. The AI Model picker in the settings panel always shows the current list.

### How many credits does the Prompt node cost?

1 credit per run for economy models such as Gemini 3.6 Flash, 2 credits for standard models such as Claude Sonnet 5, and 6 credits for premium models such as Claude Opus 5. The Very high and Max reasoning levels, and Advanced mode, bill one tier higher.

### Can the Prompt node read images, video or audio?

Yes. Connect the media to the References input. Every model reads images. Gemini 3 Flash, Gemini 3.6 Flash and Gemini 3.1 Pro also read video and audio.

### How do I get one image for each prompt the node writes?

Ask the model to separate the prompts with ===NEXT===, then connect the Items output to Generate Image. Generate Image runs once for each item.

### Why did my run fail with a "cut off" error?

The answer reached the Max Tokens limit before it was finished. The run fails, the credits are refunded, and no partial answer is passed on. Ask for a shorter answer, lower Reasoning Effort, or raise Max Tokens.

### What happened to the Generate Text and AI Agent nodes?

The Prompt node replaces both. It was called Generate Text before September 2026, and a workflow that contains an older AI Agent node opens with a Prompt node in its place.